Laboratory of Bioinformatics

Scope of research

Our main objective is to analyze, model, and simulate regulatory processes in the cell, and provide bioinformatics support for measured data. We develop computerized models of regulation of gene expression and we use statistics and artificial intelligence for analysis of proteomic and other large scale data. In proteomics we have focused on two prokaryotic microorganisms – Streptomyces and Caulobacter.

The computerized model of gene expression is based on the concept of genetic network which is in our case formalized to a principle of neural network. We assume that the rate of protein expression of each individual protein is a result of combinatorial action of all molecules which influence expression of the given gene. The influence of each molecule is expressed in a weight matrix. This principle has been expressed in a set of time dependent differential equations. Their solution allows simulation of behavior of the given system over time. The concept was successfully tested on the simple regulatory network of the phage lambda. Inverse problem - reconstruction of the weight matrix from experimentally measured time series is being solved. The reconstructed weight matrix then allows drawing of mutual interaction maps among member molecules of the system and the design of regulatory pathways.

The image of current state of gene expression can be monitored mainly by transcriptomics and proteomics. We have focused on the proteomic analysis of gene expression during development of two microorganisms – Caulobacter crescentus, which serves as a model for the study of cell cycle, and streptomycetes, which are industrially important producers of antibiotics with complex developmental cycle. We have been analyzing quantitative changes in the proteome during different stages of development in both species, including initial transition from dormant spores to metabolically active vegetative forms in Streptomyces. We use proteomic data together with statistics and artificial intelligence for identification of global trends in gene expression, principles of transition between different developmental phases, and identification of proteins and protein functional groups involved in the process. We are developing tools for data mining of complex proteomic, transcriptomic and metabolic databases.

For the storage of large scale proteomics data, together with Biozentrum, Univ. Basel, we have established a database and proteomics server SWICZ (http://proteom.biomed.cas.cz). The server hosts the databases for the above mentioned species, and is equipped with graphical interface and search engines which allow retrieving relevant information. The databases are linked with other data sources as GeneBank, and KEGG metabolic pathway server.

The lab is a member of the consortium ActinoGEN of the EU 6th framwork program .